Submission 205
Associate Professor Darcy Allen and Professor Chris Berg
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Two economists, writing in a personal capacity, argue that capturing AI's productivity gains depends on firms reorganising, which regulation can constrain, so red tape reduction is AI policy. They say this does not mean abandoning regulation's purposes. They propose reinvigorated red tape reduction and LLM-based measurement of regulation, noting these measure characteristics rather than actual economic burden.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Argues that AI productivity gains depend on firms reorganising around AI, that regulation can constrain this, and that red tape reform is therefore AI policy.
Our aim in this submission is to draw attention to a critical aspect of policy that will determine whether Australians get the productivity gains from AI. Those gains depend on firms reorganising around AI capabilities, and regulation can constrain that reorganisation. Red tape reform is therefore AI policy.
Argues that licensing rules, prescriptive standards, prior approvals and overlapping regimes impose costs on firms' capacity to experiment.
Rules that fix who may supply a service, how it must be delivered, or which approvals a new activity needs impose adaptive efficiency costs including the capacity for experimentation. We use ‘red tape’ in this broad sense. It covers not only paperwork, but the licensing rules, prescriptive standards, prior approvals and overlapping regimes that limit how firms can reorganise.
Argues that new regulation can add to these costs, citing proposed mandatory employer-workforce agreements before AI adoption as an example.
New regulation can add to these costs as readily as the existing stock. Proposals to require prior agreement before workplace AI adoption illustrate the problem. For example, there have been proposals for mandatory agreements that employers would need to reach with their workforce before introducing new AI technologies.
Says characteristics of regulation do not directly measure economic burden, which also depends on enforcement and affected firms, markets and technologies.
These are characteristics of regulation, not direct measures of economic burden. Actual burden also depends on how rules are enforced and on the firms, markets and technologies they bind.
Says a count of regulatory requirements was an imperfect proxy for economic burden but created a transparent baseline and made reform commitments checkable.
The count was an imperfect proxy for economic burden, and a fall in requirements does not mean burden fell by the same amount. Yet the measure created a transparent baseline and made commitments checkable and kept reform going after the first target was met.
Argues that capturing AI gains requires cutting existing rules and stopping new ones that hold back adaptation, while stating this does not mean abandoning the purposes of regulation.
Capturing the gains from AI therefore requires both cutting the stock of rules that holds back adaptation, and stopping new rules that add to it. This does not mean abandoning the purposes of regulation, but pursuing them without dictating how firms must achieve them.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Reinvigorate red tape reduction policy to remove barriers to business entry, new business models and resource movement, by expanding red tape reduction policies, including new measurements and mechanisms.
Reinvigorate red tape reduction policy. The Government should seek to identify and remove barriers to business entry, new business models and the movement of resources to more productive uses. It should do so by expanding red tape reduction policies, including new measurements and mechanisms.
Develop new regulatory analytics, including using large language models to measure characteristics of regulation that contribute to its burden, such as complexity, discretion and number of approvals.
The Government should develop new regulatory analytics, including using large language models (LLMs) to measure the characteristics of regulation that contribute to its burden, such as complexity, discretion and the number of approvals a new activity needs.
Use those measures to establish a baseline of the characteristics of the Commonwealth regulatory stock likely to contribute to regulatory burden.
The Government should use those measures to establish a baseline of the characteristics of the Commonwealth regulatory stock that are likely to contribute to regulatory burden.
Build Recommendation 2’s regulatory measures into repeal days, regulatory budgets and offset rules, and assess newly viable mechanisms.
included red tape repeal days, regulatory budgets and offset rules such as 1-in-x-out rules. The Government should build the measures in Recommendation 2 into these mechanisms, and assess what new mechanisms they make viable.